Updated update_copyright to work with git/hg/bzr.
authorW. Trevor King <wking@drexel.edu>
Wed, 23 Jun 2010 01:11:52 +0000 (21:11 -0400)
committerW. Trevor King <wking@drexel.edu>
Thu, 9 Feb 2012 18:04:58 +0000 (13:04 -0500)
I'd added the extra VCS support when I took it over to Hooke.

update_copyright.py

index 152829e..a600eac 100755 (executable)
 #!/usr/bin/python
 #
-# Copyright (C) 2009 W. Trevor King <wking@drexel.edu>
+# Copyright (C) 2009-2010 W. Trevor King <wking@drexel.edu>
 #
-# This program is free software; you can redistribute it and/or modify
-# it under the terms of the GNU General Public License as published by
-# the Free Software Foundation; either version 2 of the License, or
-# (at your option) any later version.
+# This file is part of Bugs Everywhere.
 #
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-# GNU General Public License for more details.
+# Bugs Everywhere is free software; you can redistribute it and/or modify it
+# under the terms of the GNU General Public License as published by the
+# Free Software Foundation, either version 2 of the License, or (at your
+# option) any later version.
 #
-# You should have received a copy of the GNU General Public License along
-# with this program; if not, write to the Free Software Foundation, Inc.,
-# 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+# Bugs Everywhere is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with Bugs Everywhere.  If not, see <http://www.gnu.org/licenses/>.
+
+"""Automatically update copyright boilerplate.
+
+This script is adapted from one written for `Bugs Everywhere`_. and
+later modified for `Hooke`_ before returning to `Bugs Everywhere`_.
+
+.. _Bugs Everywhere: http://bugseverywhere.org/
+.. _Hooke: http://code.google.com/p/hooke/
+"""
 
+import difflib
+import email.utils
+import os
 import os.path
 import re
 import sys
 import time
 
-import os
-import sys
-import select
-from threading import Thread
 
-from libbe.util.subproc import Pipe
+PROJECT_INFO = {
+    'project': 'Bugs Everywhere',
+    'vcs': 'Git',
+    }
 
-COPYRIGHT_TEXT="""#
-# This program is free software; you can redistribute it and/or modify
-# it under the terms of the GNU General Public License as published by
-# the Free Software Foundation; either version 2 of the License, or
-# (at your option) any later version.
-#
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-# GNU General Public License for more details.
-#
-# You should have received a copy of the GNU General Public License along
-# with this program; if not, write to the Free Software Foundation, Inc.,
-# 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA."""
-
-COPYRIGHT_TAG='-xyz-COPYRIGHT-zyx-' # unlikely to occur in the wild :p
-
-ALIASES = [
-    ['Ben Finney <benf@cybersource.com.au>',
-     'Ben Finney <ben+python@benfinney.id.au>',
-     'John Doe <jdoe@example.com>'],
-    ['Chris Ball <cjb@laptop.org>',
-     'Chris Ball <cjb@thunk.printf.net>'],
-    ['Gianluca Montecchi <gian@grys.it>',
-     'gian <gian@li82-39>',
-     'gianluca <gian@galactica>'],
-    ['W. Trevor King <wking@drexel.edu>',
-     'wking <wking@mjolnir>'],
-    [None,
-     'j^ <j@oil21.org>'],
-    ]
-COPYRIGHT_ALIASES = [
-    ['Aaron Bentley and Panometrics, Inc.',
-     'Aaron Bentley <abentley@panoramicfeedback.com>'],
-    ]
-EXCLUDES = [
-    ['Aaron Bentley and Panometrics, Inc.',
-     'Aaron Bentley <aaron.bentley@utoronto.ca>',]
-    ]
-
-
-IGNORED_PATHS = ['./.be/', './.bzr/', './build/']
-IGNORED_FILES = ['COPYING', 'update_copyright.py', 'catmutt']
+# Break "copyright" into "copy" and "right" to avoid matching the
+# REGEXP.
+COPY_RIGHT_TEXT="""
+This file is part of %(project)s.
 
-def _strip_email(*args):
+%(project)s is free software; you can redistribute it and/or modify it
+under the terms of the GNU General Public License as published by the
+Free Software Foundation, either version 2 of the License, or (at your
+option) any later version.
+
+%(project)s is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+General Public License for more details.
+
+You should have received a copy of the GNU General Public License
+along with %(project)s.  If not, see <http://www.gnu.org/licenses/>.
+""".strip()
+
+COPY_RIGHT_TAG='-xyz-COPY' + '-RIGHT-zyx-' # unlikely to occur in the wild :p
+
+# Convert author names to canonical forms.
+# ALIASES[<canonical name>] = <list of aliases>
+# for example,
+# ALIASES = {
+#     'John Doe <jdoe@a.com>':
+#         ['John Doe', 'jdoe', 'J. Doe <j@doe.net>'],
+#     }
+# Git-based projects are encouraged to use .mailmap instead of
+# ALIASES.  See git-shortlog(1) for details.
+ALIASES = {
+    'Aaron Bentley <abentley@panoramicfeedback.com>':
+        ['Aaron Bentley <aaron.bentley@utoronto.ca>'],
+    'Panometrics, Inc.':
+        ['Aaron Bentley and Panometrics, Inc.'],
+    'Ben Finney <benf@cybersource.com.au>':
+        ['Ben Finney <ben+python@benfinney.id.au>',
+         'John Doe <jdoe@example.com>'],
+    'Chris Ball <cjb@laptop.org>':
+        ['Chris Ball <cjb@thunk.printf.net>'],
+    'Gianluca Montecchi <gian@grys.it>':
+        ['gian <gian@li82-39>',
+         'gianluca <gian@galactica>'],
+    'W. Trevor King <wking@drexel.edu>':
+        ['wking <wking@mjolnir>',
+         'wking <wking@thialfi>'],
+    None:
+        ['j^ <j@oil21.org>'],
+    }
+
+# List of paths that should not be scanned for copyright updates.
+# IGNORED_PATHS = ['./.git/']
+IGNORED_PATHS = ['./.be/', './.git/', './build/', './doc/.build/']
+# List of files that should not be scanned for copyright updates.
+# IGNORED_FILES = ['COPYING']
+IGNORED_FILES = ['COPYING', 'catmutt']
+
+# Work around missing author holes in the VCS history.
+# AUTHOR_HACKS[<path tuple>] = [<missing authors]
+# for example, if John Doe contributed to module.py but wasn't listed
+# in the VCS history of that file:
+# AUTHOR_HACKS = {
+#     ('path', 'to', 'module.py'):['John Doe'],
+#     }
+AUTHOR_HACKS = {}
+
+# Work around missing year holes in the VCS history.
+# YEAR_HACKS[<path tuple>] = <original year>
+# for example, if module.py was published in 2008 but the VCS history
+# only goes back to 2010:
+# YEAR_HACKS = {
+#     ('path', 'to', 'module.py'):2008,
+#     }
+YEAR_HACKS = {}
+
+# Helpers for VCS-specific commands
+
+def splitpath(path):
+    """Recursively split a path into elements.
+
+    Examples
+    --------
+
+    >>> splitpath(os.path.join('a', 'b', 'c'))
+    ('a', 'b', 'c')
+    >>> splitpath(os.path.join('.', 'a', 'b', 'c'))
+    ('a', 'b', 'c')
     """
+    path = os.path.normpath(path)
+    elements = []
+    while True:
+        dirname,basename = os.path.split(path)
+        elements.insert(0,basename)
+        if dirname in ['', '.']:
+            break
+        path = dirname
+    return tuple(elements)
+
+# VCS-specific commands
+
+if PROJECT_INFO['vcs'] == 'Git':
+
+    import subprocess
+
+    _MSWINDOWS = sys.platform == 'win32'
+    _POSIX = not _MSWINDOWS
+
+    def invoke(args, stdin=None, stdout=subprocess.PIPE, stderr=subprocess.PIPE, expect=(0,)):
+        """
+        expect should be a tuple of allowed exit codes.
+        """
+        try :
+            if _POSIX:
+                q = subprocess.Popen(args, stdin=subprocess.PIPE,
+                                     stdout=stdout, stderr=stderr)
+            else:
+                assert _MSWINDOWS == True, 'invalid platform'
+                # win32 don't have os.execvp() so run the command in a shell
+                q = subprocess.Popen(args, stdin=subprocess.PIPE,
+                                     stdout=stdout, stderr=stderr, shell=True)
+        except OSError, e:
+            raise ValueError([args, e])
+        stdout,stderr = q.communicate(input=stdin)
+        status = q.wait()
+        if status not in expect:
+            raise ValueError([args, status, stdout, stderr])
+        return status, stdout, stderr
+
+    def git_cmd(*args):
+        status,stdout,stderr = invoke(['git'] + list(args))
+        return stdout.rstrip('\n')
+
+    version = git_cmd('--version').split(' ')[-1]
+    if version.startswith('1.5.'):
+        # Author name <author email>
+        author_format = '--pretty=format:%an <%ae>'
+        year_format = ['--pretty=format:%ai']  # Author date
+        # YYYY-MM-DD HH:MM:SS Z
+        # Earlier versions of Git don't seem to recognize --date=short
+    else:
+        author_format = '--pretty=format:%aN <%aE>'
+        year_format = ['--pretty=format:%ad',  # Author date
+                       '--date=short']         # YYYY-MM-DD
+
+    def original_year(filename, year_hacks=YEAR_HACKS):
+        output = git_cmd(*(['log', '--follow']
+                           + year_format
+                           + [filename]))
+        if version.startswith('1.5.'):
+            output = '\n'.join([x.split()[0] for x in output.splitlines()])
+        years = [int(line.split('-', 1)[0]) for line in output.splitlines()]
+        if splitpath(filename) in year_hacks:
+            years.append(year_hacks[splitpath(filename)])
+        years.sort()
+        return years[0]
+
+    def authors(filename, author_hacks=AUTHOR_HACKS):
+        output = git_cmd('log', '--follow', author_format,
+                         filename)
+        ret = list(set(output.splitlines()))
+        if splitpath(filename) in author_hacks:
+            ret.extend(author_hacks[splitpath(filename)])
+        return ret
+
+    def authors_list(author_hacks=AUTHOR_HACKS):
+        output = git_cmd('log', author_format)
+        ret = list(set(output.splitlines()))
+        for path,authors in author_hacks.items():
+            ret.extend(authors)
+        return ret
+
+    def is_versioned(filename):
+        output = git_cmd('log', '--follow', filename)
+        if len(output) == 0:
+            return False
+        return True
+
+elif PROJECT_INFO['vcs'] == 'Mercurial':
+
+    import StringIO
+    import mercurial
+    import mercurial.dispatch
+
+    def mercurial_cmd(*args):
+        cwd = os.getcwd()
+        stdout = sys.stdout
+        stderr = sys.stderr
+        tmp_stdout = StringIO.StringIO()
+        tmp_stderr = StringIO.StringIO()
+        sys.stdout = tmp_stdout
+        sys.stderr = tmp_stderr
+        try:
+            mercurial.dispatch.dispatch(list(args))
+        finally:
+            os.chdir(cwd)
+            sys.stdout = stdout
+            sys.stderr = stderr
+        return (tmp_stdout.getvalue().rstrip('\n'),
+                tmp_stderr.getvalue().rstrip('\n'))
+
+    def original_year(filename, year_hacks=YEAR_HACKS):
+        # shortdate filter: YEAR-MONTH-DAY
+        output,error = mercurial_cmd('log', '--follow',
+                                     '--template', '{date|shortdate}\n',
+                                     filename)
+        years = [int(line.split('-', 1)[0]) for line in output.splitlines()]
+        if splitpath(filename) in year_hacks:
+            years.append(year_hacks[splitpath(filename)])
+        years.sort()
+        return years[0]
+
+    def authors(filename, author_hacks=AUTHOR_HACKS):
+        output,error = mercurial_cmd('log', '--follow',
+                                     '--template', '{author}\n',
+                                     filename)
+        ret = list(set(output.splitlines()))
+        if splitpath(filename) in author_hacks:
+            ret.extend(author_hacks[splitpath(filename)])
+        return ret
+
+    def authors_list(author_hacks=AUTHOR_HACKS):
+        output,error = mercurial_cmd('log', '--template', '{author}\n')
+        ret = list(set(output.splitlines()))
+        for path,authors in author_hacks.items():
+            ret.extend(authors)
+        return ret
+
+    def is_versioned(filename):
+        output,error = mercurial_cmd('log', '--follow', filename)
+        if len(error) > 0:
+            return False
+        return True
+
+elif PROJECT_INFO['vcs'] == 'Bazaar':
+
+    import StringIO
+    import bzrlib
+    import bzrlib.builtins
+    import bzrlib.log
+
+    class LogFormatter (bzrlib.log.LogFormatter):
+        supports_merge_revisions = True
+        preferred_levels = 0
+        supports_deta = False
+        supports_tags = False
+        supports_diff = False
+
+        def log_revision(self, revision):
+            raise NotImplementedError
+
+    class YearLogFormatter (LogFormatter):
+        def log_revision(self, revision):
+            self.to_file.write(
+                time.strftime('%Y', time.gmtime(revision.rev.timestamp))
+                +'\n')
+
+    class AuthorLogFormatter (LogFormatter):
+        def log_revision(self, revision):
+            authors = revision.rev.get_apparent_authors()
+            self.to_file.write('\n'.join(authors)+'\n')
+
+    def original_year(filename, year_hacks=YEAR_HACKS):
+        cmd = bzrlib.builtins.cmd_log()
+        cmd.outf = StringIO.StringIO()
+        cmd.run(file_list=[filename], log_format=YearLogFormatter, levels=0)
+        years = [int(year) for year in set(cmd.outf.getvalue().splitlines())]
+        if splitpath(filename) in year_hacks:
+            years.append(year_hacks[splitpath(filename)])
+        years.sort()
+        return years[0]
+
+    def authors(filename, author_hacks=AUTHOR_HACKS):
+        cmd = bzrlib.builtins.cmd_log()
+        cmd.outf = StringIO.StringIO()
+        cmd.run(file_list=[filename], log_format=AuthorLogFormatter, levels=0)
+        ret = list(set(cmd.outf.getvalue().splitlines()))
+        if splitpath(filename) in author_hacks:
+            ret.extend(author_hacks[splitpath(filename)])
+        return ret
+
+    def authors_list(author_hacks=AUTHOR_HACKS):
+        cmd = bzrlib.builtins.cmd_log()
+        cmd.outf = StringIO.StringIO()
+        cmd.run(log_format=AuthorLogFormatter, levels=0)
+        output = cmd.outf.getvalue()
+        ret = list(set(cmd.outf.getvalue().splitlines()))
+        for path,authors in author_hacks.items():
+            ret.extend(authors)
+        return ret
+
+    def is_versioned(filename):
+        cmd = bzrlib.builtins.cmd_log()
+        cmd.outf = StringIO.StringIO()
+        cmd.run(file_list=[filename])
+        return True
+
+else:
+    raise NotImplementedError('Unrecognized VCS: %(vcs)s' % PROJECT_INFO)
+
+# General utility commands
+
+def _strip_email(*args):
+    """Remove email addresses from a series of names.
+
+    Examples
+    --------
+
     >>> _strip_email('J Doe <jdoe@a.com>')
     ['J Doe']
     >>> _strip_email('J Doe <jdoe@a.com>', 'JJJ Smith <jjjs@a.com>')
@@ -83,127 +356,109 @@ def _strip_email(*args):
     for i,arg in enumerate(args):
         if arg == None:
             continue
-        index = arg.find('<')
-        if index > 0:
-            args[i] = arg[:index].rstrip()
+        author,addr = email.utils.parseaddr(arg)
+        args[i] = author
     return args
 
-def _replace_aliases(authors, with_email=True, aliases=None,
-                     excludes=None):
+def _reverse_aliases(aliases):
+    """Reverse an `aliases` dict.
+
+    Input:   key: canonical name,  value: list of aliases
+    Output:  key: alias,           value: canonical name
+
+    Examples
+    --------
+
+    >>> aliases = {
+    ...     'J Doe <jdoe@a.com>':['Johnny <jdoe@b.edu>', 'J'],
+    ...     'JJJ Smith <jjjs@a.com>':['Jingly <jjjs@b.edu>'],
+    ...     None:['Anonymous <a@a.com>'],
+    ...     }
+    >>> r = _reverse_aliases(aliases)
+    >>> for item in sorted(r.items()):
+    ...     print item
+    ('Anonymous <a@a.com>', None)
+    ('J', 'J Doe <jdoe@a.com>')
+    ('Jingly <jjjs@b.edu>', 'JJJ Smith <jjjs@a.com>')
+    ('Johnny <jdoe@b.edu>', 'J Doe <jdoe@a.com>')
     """
-    >>> aliases = [['J Doe and C, Inc.', 'J Doe <jdoe@c.com>'],
-    ...            ['J Doe <jdoe@a.com>', 'Johnny <jdoe@b.edu>'],
-    ...            ['JJJ Smith <jjjs@a.com>', 'Jingly <jjjs@b.edu>'],
-    ...            [None, 'Anonymous <a@a.com>']]
-    >>> excludes = [['J Doe and C, Inc.', 'J Doe <jdoe@a.com>']]
+    output = {}
+    for canonical_name,_aliases in aliases.items():
+        for alias in _aliases:
+            output[alias] = canonical_name
+    return output
+
+def _replace_aliases(authors, with_email=True, aliases=None):
+    """Consolidate and sort `authors`.
+
+    Make the replacements listed in the `aliases` dict (key: canonical
+    name, value: list of aliases).  If `aliases` is ``None``, default
+    to ``ALIASES``.
+
+    >>> aliases = {
+    ...     'J Doe <jdoe@a.com>':['Johnny <jdoe@b.edu>'],
+    ...     'JJJ Smith <jjjs@a.com>':['Jingly <jjjs@b.edu>'],
+    ...     None:['Anonymous <a@a.com>'],
+    ...     }
     >>> _replace_aliases(['JJJ Smith <jjjs@a.com>', 'Johnny <jdoe@b.edu>',
     ...                   'Jingly <jjjs@b.edu>', 'Anonymous <a@a.com>'],
-    ...                  with_email=True, aliases=aliases, excludes=excludes)
+    ...                  with_email=True, aliases=aliases)
     ['J Doe <jdoe@a.com>', 'JJJ Smith <jjjs@a.com>']
     >>> _replace_aliases(['JJJ Smith', 'Johnny', 'Jingly', 'Anonymous'],
-    ...                  with_email=False, aliases=aliases, excludes=excludes)
+    ...                  with_email=False, aliases=aliases)
     ['J Doe', 'JJJ Smith']
     >>> _replace_aliases(['JJJ Smith <jjjs@a.com>', 'Johnny <jdoe@b.edu>',
-    ...                   'Jingly <jjjs@b.edu>', 'J Doe <jdoe@c.com>'],
-    ...                  with_email=True, aliases=aliases, excludes=excludes)
-    ['J Doe and C, Inc.', 'JJJ Smith <jjjs@a.com>']
+    ...                   'Jingly <jjjs@b.edu>', 'J Doe <jdoe@a.com>'],
+    ...                  with_email=True, aliases=aliases)
+    ['J Doe <jdoe@a.com>', 'JJJ Smith <jjjs@a.com>']
     """
     if aliases == None:
         aliases = ALIASES
-    if excludes == None:
-        excludes = EXCLUDES
     if with_email == False:
-        aliases = [_strip_email(*alias) for alias in aliases]
-        exclude = [_strip_email(*exclude) for exclude in excludes]
+        aliases = dict([(_strip_email(author)[0], _strip_email(*_aliases))
+                        for author,_aliases in aliases.items()])
+    rev_aliases = _reverse_aliases(aliases)
     for i,author in enumerate(authors):
-        for alias in aliases:
-            if author in alias[1:]:
-                authors[i] = alias[0]
-                break
-    for i,author in enumerate(authors):
-        for exclude in excludes:
-            if author in exclude[1:] and exclude[0] in authors:
-                authors[i] = None
-    authors = sorted(set(authors))
+        if author in rev_aliases:
+            authors[i] = rev_aliases[author]
+    authors = sorted(list(set(authors)))
     if None in authors:
         authors.remove(None)
     return authors
 
-def authors_list():
-    p = Pipe([['bzr', 'log', '-n0'],
-              ['grep', '^ *committer\|^ *author'],
-              ['cut', '-d:', '-f2'],
-              ['sed', 's/ <.*//;s/^ *//'],
-              ['sort'],
-              ['uniq']])
-    assert p.status == 0, p.statuses
-    authors = p.stdout.rstrip().split('\n')
-    return _replace_aliases(authors, with_email=False)
-
-def update_authors(verbose=True):
-    print "updating AUTHORS"
-    f = file('AUTHORS', 'w')
-    authors_text = 'Bugs Everywhere was written by:\n%s\n' % '\n'.join(authors_list())
-    f.write(authors_text)
-    f.close()
-
-def ignored_file(filename, ignored_paths=None, ignored_files=None):
-    """
-    >>> ignored_paths = ['./a/', './b/']
-    >>> ignored_files = ['x', 'y']
-    >>> ignored_file('./a/z', ignored_paths, ignored_files)
-    True
-    >>> ignored_file('./ab/z', ignored_paths, ignored_files)
-    False
-    >>> ignored_file('./ab/x', ignored_paths, ignored_files)
-    True
-    >>> ignored_file('./ab/xy', ignored_paths, ignored_files)
-    False
-    >>> ignored_file('./z', ignored_paths, ignored_files)
-    False
-    """
-    if ignored_paths == None:
-        ignored_paths = IGNORED_PATHS
-    if ignored_files == None:
-        ignored_files = IGNORED_FILES
-    for path in ignored_paths:
-        if filename.startswith(path):
-            return True
-    if os.path.basename(filename) in ignored_files:
-        return True
-    if os.path.abspath(filename) != os.path.realpath(filename):
-        return True # symink somewhere in path...
-    return False
-
-def _copyright_string(orig_year, final_year, authors):
+def _copyright_string(original_year, final_year, authors, prefix=''):
     """
-    >>> print _copyright_string(orig_year=2005,
+    >>> print _copyright_string(original_year=2005,
     ...                         final_year=2005,
-    ...                         authors=['A <a@a.com>', 'B <b@b.edu>']
+    ...                         authors=['A <a@a.com>', 'B <b@b.edu>'],
+    ...                         prefix='# '
     ...                        ) # doctest: +ELLIPSIS
     # Copyright (C) 2005 A <a@a.com>
     #                    B <b@b.edu>
     #
-    # This program...
-    >>> print _copyright_string(orig_year=2005,
+    # This file...
+    >>> print _copyright_string(original_year=2005,
     ...                         final_year=2009,
     ...                         authors=['A <a@a.com>', 'B <b@b.edu>']
     ...                        ) # doctest: +ELLIPSIS
-    Copyright (C) 2005-2009 A <a@a.com>
-                            B <b@b.edu>
-    #
-    # This program...
+    Copyright (C) 2005-2009 A <a@a.com>
+                            B <b@b.edu>
+    <BLANKLINE>
+    This file...
     """
-    if orig_year == final_year:
-        date_range = '%s' % orig_year
+    if original_year == final_year:
+        date_range = '%s' % original_year
     else:
-        date_range = '%s-%s' % (orig_year, final_year)
-    lines = ['Copyright (C) %s %s' % (date_range, authors[0])]
+        date_range = '%s-%s' % (original_year, final_year)
+    lines = ['Copyright (C) %s %s' % (date_range, authors[0])]
     for author in authors[1:]:
-        lines.append('#' +
-                     ' '*(len(' Copyright (C) ')+len(date_range)+1) +
+        lines.append(' '*(len('Copyright (C) ')+len(date_range)+1) +
                      author)
-    return '%s\n%s' % ('\n'.join(lines), COPYRIGHT_TEXT)
+    lines.append('')
+    lines.extend((COPY_RIGHT_TEXT % PROJECT_INFO).splitlines())
+    for i,line in enumerate(lines):
+        lines[i] = (prefix + line).rstrip()
+    return '\n'.join(lines)
 
 def _tag_copyright(contents):
     """
@@ -215,75 +470,149 @@ def _tag_copyright(contents):
     ... (copyright ends)
     ... bla bla bla
     ... '''
-    >>> print _tag_copyright(contents),
+    >>> print _tag_copyright(contents).replace('COPY-RIGHT', 'CR')
     Some file
     bla bla
-    -xyz-COPYRIGHT-zyx-
+    -xyz-CR-zyx-
     (copyright ends)
     bla bla bla
+    <BLANKLINE>
     """
     lines = []
     incopy = False
     for line in contents.splitlines():
         if incopy == False and line.startswith('# Copyright'):
             incopy = True
-            lines.append(COPYRIGHT_TAG)
+            lines.append(COPY_RIGHT_TAG)
         elif incopy == True and not line.startswith('#'):
             incopy = False
         if incopy == False:
             lines.append(line.rstrip('\n'))
     return '\n'.join(lines)+'\n'
 
-def _update_copyright(contents, orig_year, authors):
+def _update_copyright(contents, original_year, authors):
+    """
+    >>> contents = '''Some file
+    ... bla bla
+    ... # Copyright (copyright begins)
+    ... # (copyright continues)
+    ... # bla bla bla
+    ... (copyright ends)
+    ... bla bla bla
+    ... '''
+    >>> print _update_copyright(contents, 2008, ['Jack', 'Jill']
+    ...     ) # doctest: +ELLIPSIS, +REPORT_UDIFF
+    Some file
+    bla bla
+    # Copyright (C) 2008-... Jack
+    #                         Jill
+    #
+    # This file...
+    (copyright ends)
+    bla bla bla
+    <BLANKLINE>
+    """
     current_year = time.gmtime()[0]
-    copyright_string = _copyright_string(orig_year, current_year, authors)
+    copyright_string = _copyright_string(
+        original_year, current_year, authors, prefix='# ')
     contents = _tag_copyright(contents)
-    return contents.replace(COPYRIGHT_TAG, copyright_string)
-
-def update_file(filename, verbose=True):
-    if verbose == True:
-        print "updating", filename
-    contents = file(filename, 'r').read()
-
-    p = Pipe([['bzr', 'log', '-n0', filename],
-              ['grep', '^ *timestamp: '],
-              ['tail', '-n1'],
-              ['sed', 's/^ *//;'],
-              ['cut', '-b', '16-19']])
-    if p.status != 0:
-        assert p.statuses[0] == 3, p.statuses
-        return # bzr doesn't version that file
-    assert p.status == 0, p.statuses
-    orig_year = int(p.stdout.strip())
-
-    p = Pipe([['bzr', 'log', '-n0', filename],
-              ['grep', '^ *author: \|^ *committer: '],
-              ['cut', '-d:', '-f2'],
-              ['sed', 's/^ *//;s/ *$//'],
-              ['sort'],
-              ['uniq']])
-    assert p.status == 0, p.statuses
-    authors = p.stdout.rstrip().split('\n')
-    authors = _replace_aliases(authors, with_email=True,
-                               aliases=ALIASES+COPYRIGHT_ALIASES)
-
-    contents = _update_copyright(contents, orig_year, authors)
-    f = file(filename, 'w')
-    f.write(contents)
+    return contents.replace(COPY_RIGHT_TAG, copyright_string)
+
+def ignored_file(filename, ignored_paths=None, ignored_files=None,
+                 check_disk=True, check_vcs=True):
+    """
+    >>> ignored_paths = ['./a/', './b/']
+    >>> ignored_files = ['x', 'y']
+    >>> ignored_file('./a/z', ignored_paths, ignored_files, False, False)
+    True
+    >>> ignored_file('./ab/z', ignored_paths, ignored_files, False, False)
+    False
+    >>> ignored_file('./ab/x', ignored_paths, ignored_files, False, False)
+    True
+    >>> ignored_file('./ab/xy', ignored_paths, ignored_files, False, False)
+    False
+    >>> ignored_file('./z', ignored_paths, ignored_files, False, False)
+    False
+    """
+    if ignored_paths == None:
+        ignored_paths = IGNORED_PATHS
+    if ignored_files == None:
+        ignored_files = IGNORED_FILES
+    if check_disk == True and os.path.isfile(filename) == False:
+        return True
+    for path in ignored_paths:
+        if filename.startswith(path):
+            return True
+    if os.path.basename(filename) in ignored_files:
+        return True
+    if check_vcs == True and is_versioned(filename) == False:
+        return True
+    return False
+
+def _set_contents(filename, contents, original_contents=None, dry_run=False,
+                  verbose=0):
+    if original_contents == None and os.path.isfile(filename):
+        f = open(filename, 'r')
+        original_contents = f.read()
+        f.close()
+    if verbose > 0:
+        print "checking %s ... " % filename,
+    if contents != original_contents:
+        if verbose > 0:
+            if original_contents == None:
+                print "[creating]"
+            else:
+                print "[updating]"
+        if verbose > 1 and original_contents != None:
+            print '\n'.join(
+                difflib.unified_diff(
+                    original_contents.splitlines(), contents.splitlines(),
+                    fromfile=os.path.normpath(os.path.join('a', filename)),
+                    tofile=os.path.normpath(os.path.join('b', filename)),
+                    n=3, lineterm=''))
+        if dry_run == False:
+            f = file(filename, 'w')
+            f.write(contents)
+            f.close()
+    elif verbose > 0:
+        print "[no change]"
+
+# Update commands
+
+def update_authors(authors_fn=authors_list, dry_run=False, verbose=0):
+    authors = authors_fn()
+    authors = _replace_aliases(authors, with_email=True, aliases=ALIASES)
+    new_contents = '%s was written by:\n%s\n' % (
+        PROJECT_INFO['project'],
+        '\n'.join(authors)
+        )
+    _set_contents('AUTHORS', new_contents, dry_run=dry_run, verbose=verbose)
+
+def update_file(filename, original_year_fn=original_year, authors_fn=authors,
+                dry_run=False, verbose=0):
+    f = file(filename, 'r')
+    contents = f.read()
     f.close()
 
-def update_files(files=None):
+    original_year = original_year_fn(filename)
+    authors = authors_fn(filename)
+    authors = _replace_aliases(authors, with_email=True, aliases=ALIASES)
+
+    new_contents = _update_copyright(contents, original_year, authors)
+    _set_contents(filename, contents=new_contents, original_contents=contents,
+                  dry_run=dry_run, verbose=verbose)
+
+def update_files(files=None, dry_run=False, verbose=0):
     if files == None or len(files) == 0:
-        p = Pipe([['grep', '-rc', '# Copyright', '.'],
-                  ['grep', '-v', ':0$'],
-                  ['cut', '-d:', '-f1']])
-        assert p.status == 0
-        files = p.stdout.rstrip().split('\n')
+        files = []
+        for dirpath,dirnames,filenames in os.walk('.'):
+            for filename in filenames:
+                files.append(os.path.join(dirpath, filename))
 
     for filename in files:
         if ignored_file(filename) == True:
             continue
-        update_file(filename)
+        update_file(filename, dry_run=dry_run, verbose=verbose)
 
 def test():
     import doctest
@@ -291,28 +620,34 @@ def test():
 
 if __name__ == '__main__':
     import optparse
-    usage = """%prog [options] [file ...]
+    import sys
+
+    usage = """%%prog [options] [file ...]
 
 Update copyright information in source code with information from
-the bzr repository.  Run from the BE repository root.
+the %(vcs)s repository.  Run from the %(project)s repository root.
 
 Replaces every line starting with '^# Copyright' and continuing with
 '^#' with an auto-generated copyright blurb.  If you want to add
 #-commented material after a copyright blurb, please insert a blank
-line between the blurb and your comment (as in this file), so the
-next run of update_copyright.py doesn't clobber your comment.
+line between the blurb and your comment, so the next run of
+``update_copyright.py`` doesn't clobber your comment.
 
 If no files are given, a list of files to update is generated
 automatically.
-"""
+""" % PROJECT_INFO
     p = optparse.OptionParser(usage)
     p.add_option('--test', dest='test', default=False,
                  action='store_true', help='Run internal tests and exit')
+    p.add_option('--dry-run', dest='dry_run', default=False,
+                 action='store_true', help="Don't make any changes")
+    p.add_option('-v', '--verbose', dest='verbose', default=0,
+                 action='count', help='Increment verbosity')
     options,args = p.parse_args()
 
     if options.test == True:
         test()
         sys.exit(0)
 
-    update_authors()
-    update_files(files=args)
+    update_authors(dry_run=options.dry_run, verbose=options.verbose)
+    update_files(files=args, dry_run=options.dry_run, verbose=options.verbose)